A model-based economic analysis of pre-pandemic …smr.csse.uwa.edu.au/pdf/2014 Halder A model-based economic analy… · A model-based economic analysis of pre-pandemic influenza

Halder et al. BMC Infectious Diseases 2014, 14:266http://www.biomedcentral.com/1471-2334/14/266

RESEARCH ARTICLE Open Access

A model-based economic analysis of pre-pandemicinfluenza vaccination cost-effectivenessNilimesh Halder, Joel K Kelso* and George J Milne

Abstract

Background: A vaccine matched to a newly emerged pandemic influenza virus would require a production time of atleast 6 months with current proven techniques, and so could only be used reactively after the peak of the pandemic. Apre-pandemic vaccine, although probably having lower efficacy, could be produced and used pre-emptively. Whileseveral previous studies have investigated the cost effectiveness of pre-emptive vaccination strategies, they have notbeen directly compared to realistic reactive vaccination strategies.

Methods: An individual-based simulation model of ~30,000 people was used to examine a pre-emptive vaccinationstrategy, assuming vaccination conducted prior to a pandemic using a low-efficacy vaccine. A reactive vaccinationstrategy, assuming a 6-month delay between pandemic emergence and availability of a high-efficacy vaccine, wasalso modelled. Social distancing and antiviral interventions were examined in combination with these alternativevaccination strategies. Moderate and severe pandemics were examined, based on estimates of transmissibility andclinical severity of the 1957 and 1918 pandemics respectively, and the cost effectiveness of each strategy was evaluated.

Results: Provided that a pre-pandemic vaccine achieved at least 30% efficacy, pre-emptive vaccination strategies werefound to be more cost effective when compared to reactive vaccination strategies. Reactive vaccination coupledwith sustained social distancing and antiviral interventions was found to be as effective at saving lives as pre-emptivevaccination coupled with limited duration social distancing and antiviral use, with both strategies saving approximately420 life-years per 10,000 population for a moderate pandemic with a basic reproduction number of 1.9 and case fatalityrate of 0.25%. Reactive vaccination was however more costly due to larger productivity losses incurred by sustainedsocial distancing, costing $8 million per 10,000 population ($19,074/LYS) versus $6.8 million per 10,000 population($15,897/LYS) for a pre-emptive vaccination strategy. Similar trends were observed for severe pandemics.

Conclusions: Compared to reactive vaccination, pre-emptive strategies would be more effective and more costeffective, conditional on the pre-pandemic vaccine being able to achieve a certain level of coverage and efficacy.Reactive vaccination strategies exist which are as effective at mortality reduction as pre-emptive strategies, thoughthey are less cost effective.

Keywords: Pre-emptive vaccination, Reactive vaccination, Pandemic influenza, Cost effectiveness,Simulation model

* Correspondence: [email protected] of Computer Science and Software Engineering, The University ofWestern Australia, 35 Stirling Highway, Crawley, WA 6009, Australia

© 2014 Halder et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly credited. The Creative Commons Public DomainDedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article,unless otherwise stated.

mailto:[email protected]

http://creativecommons.org/licenses/by/2.0

http://creativecommons.org/publicdomain/zero/1.0/

Halder et al. BMC Infectious Diseases 2014, 14:266 Page 2 of 19http://www.biomedcentral.com/1471-2334/14/266

BackgroundThe threat of an influenza pandemic due to a novel virusstrain or subtype to which humans have little or noimmunity is a major public health concern. The lastfour influenza pandemics have ranged in severity frommild, having an estimated case fatality ratio (CFR) between0.01% and 0.08% [1] as in the case of the 2009 pandemic,to severe, having an estimated case fatality ratio between0.74% and 1.8% [2-4] as in the case of the 1918 pandemic.In recent years, avian influenza subtypes such as H5N1 andH7N9 have begun circulating in domestic bird populationsin South-East Asia and China. These subtypes result in highcase fatality rates in humans who have contracted influenzafrom infected birds, having an estimated CFR in patientsadmitted to hospital of between 30% and 70% [5-10]. Thismay lead to a major public health disaster if such a virusmutates or reassorts into a form transmissible betweenhumans. Recent research studies highlighted the dangerof H5N1 mutation into a form readily transmissiblebetween mammal species, namely ferrets [11-13], dem-onstrating potential transmissibility between humans.Health authorities of many countries, often working with

the World Health Organisation (WHO), have establishedpandemic mitigation plans that advocate the use of a rangeof intervention strategies including social distancing,neuraminidase inhibitor antiviral drugs, and vaccinationas the major defences against pandemic associated illnessand mortality.Although vaccination is an effective and long-lasting

solution to the threat posed by a pandemic influenzastrain, the likely scenario, born out in the 2009 pan-demic, is that a newly emerged influenza pandemic willhave spread to most parts of the world before a vaccinematched to the pandemic subtype can be produced. Ahigh efficacy influenza vaccine matched to the circulatingpandemic strain could take at least 6 months from virusisolation to final vaccine production with current proventechniques [14].In case of a highly transmissible and pathogenic pan-

demic, interventions that can be activated withoutdelay will be required to combat pandemic spread untila matched vaccine can be developed. Modelling studieshave indicated that rigorously deployed social distancingand antivirals measures could suppress pandemic spreaduntil a vaccine becomes available [15,16]. However thesocial disruption and economic cost of maintaining thesemeasures for many months could be very large, and mayrender them practically infeasible.An alternative (or possibly complementary) intervention

measure is a pre-pandemic vaccine: a vaccine producedfrom isolated influenza virus strains that are consideredmost likely to emerge as pandemic strains potentiallylethal to humans. Such a vaccine could be developedand used in anticipation of a pandemic, and would not

be subject to the development delay of a reactively de-veloped vaccine. In recent years, the candidate H5N1vaccines have been advocated for pre-emptive use to com-bat against a potentially lethal pandemic that could causeby H5N1 [17-27]. Modelling of pre-emptive vaccinationalso suggested that pre-pandemic vaccines could mitigatesuch pandemics [28-31].Several modelling studies [16,28-35] have examined the

cost-effectiveness of the use of vaccination as a pandemicinfluenza mitigation measure. Milne et al. [28] demon-strated the mitigating effect of vaccination with pre-pandemic H5N1 vaccines. Baguelin et al., Prosser et al.,and Lugnér et al. [32,33,35] examined the economicoutcomes of various influenza vaccination strategies;the studies of Lee et al., Newall et al., and Khazeni et al.[30,31,34] included strategies involving antiviral drugsas well as vaccination; while the studies of Andrasdóttiret al. and Sander et al.[16,29] also included varioussocial distancing strategies in their analysis. Recently,we have investigated the cost-effectiveness of reactivevaccination taking into account the probable 6-monthdelay in vaccine availability, with and without combinedsocial distancing and antiviral interventions [15]. Theeconomic outcomes and cost effectiveness analysis ofreactive vaccination in conjunction with social distancingand antiviral drugs were also investigated in [16,35].However, no previous study has explicitly addressed

the cost-effectiveness of pre-emptive vaccination strategyin highly plausible scenarios where pre-emptive vaccin-ation is compared to reactive vaccination with a 6-monthdelay, and where both vaccination strategies are con-sidered with and without rigorous social distancingand antiviral use.This modelling study addresses this knowledge gap

using an established individual-based simulation modelof a developed country [15,36-42] by simulating bothpre-emptive and reactive vaccination strategies, com-bined with a range of social distancing and antiviralmeasures, for plausible future pandemics with differentpandemic transmissibility and severity characteristics.The costs to society of each strategy and pandemic sce-nario stemming from medical treatment, pharmaceut-ical costs and productivity losses are calculated with aneconomical model, and the effectiveness of each strat-egy is assessed in terms of Life Years Saved (LYS). Thiswork continues and completes a sequence of previousstudies by the authors on pandemic influenza vaccin-ation: in [28] we assessed the potential effectivenessof pre-pandemic vaccines, but did not address cost-effectiveness or reactive vaccination strategies; whilein [15] we assessed the cost-effectiveness of reactivevaccination strategies combined with plausible socialdistancing and antiviral measures, but did not addresspre-pandemic vaccines.


MethodsIn this section we describe the simulation model andeconomic analysis used in this study, its application topre-pandemic vaccination. Readers who are familiar withindividual-based epidemic spread modelling and cost-effectiveness analysis, and the previous studies of theauthors in particular, may be most interested in the“Pre-emptive vaccination” and “Vaccination scenarios”subsections.

Simulation modelA detailed, individual-based simulation model of a com-munity in Western Australia (Albany, population ~30,000)was used to simulate the dynamics of an influenza pan-demic, under various public health mitigation strategies,and for pandemics of different transmissibility and severitycharacteristics. Data produced by the model indicating theinfectious history of each individual in the community wasused to determine health outcomes of each individual,considering symptomatic or asymptomatic infection, hos-pitalisation, ICU treatment, and death. Together withproductivity losses due to illness and the interventionsthemselves, these outcomes were used to estimate the costand cost effectiveness of alternative intervention strategies.The Albany simulation model used in this study has

been used previously in a series of studies addressingthe effectiveness and cost effectiveness of various public

Figure 1 Synopsis of methodology.

health pandemic influenza mitigation measures, includ-ing the cost effectiveness of reactive vaccination strat-egies with respect to the delay in vaccine development[15], the economic evaluation and cost effectivenessanalysis of social distancing and antiviral drug strategiesthat were used during the 2009 H1N1 pandemic [39],combinations of antiviral and different social distancinginterventions for the future pandemics with mild,moderate and severe characteristics [15,40,41], and theeffectiveness of vaccination with pre-pandemic H5N1vaccines [28]. For the purposes of this study, a model ofpre-emptive vaccination was added which allows fordifferent levels of vaccine efficacy and coverage, andincludes the ongoing cost of vaccine development anddeployment required to keep pre-emptive vaccinationcoverage current in the population. The model used inthe present study carries over all other assumptionsfrom previous studies that do not relate to pre-emptivevaccination.A full description of the model appears in Additional

file 1; model parameter values are summarised in TableA1.1 of that file. An overview of methodology used forthe study is given in Figure 1, while Table 1 describesthe terminology used and the definition of pandemiccharacteristics, together with the alternative vaccination,antiviral and social distancing interventions analysed inthe study.

Table 1 Terminology and definition

Terminology Definition

Pandemic characteristics

Moderate pandemics A moderate pandemic is defined as a pandemic with a transmissibility of R = 1.9 and case fatality rate of 0.25%.

Severe pandemics An extreme pandemic is defined as a pandemic with a transmissibility of R = 2.7 and case fatality rate of 1.5%.

Vaccinations

Pre-emptive vaccination A two-dose pre-emptive vaccination strategy is considered assuming that individuals should be vaccinated withpre-pandemic vaccine annually. This pre-pandemic vaccine should be updated in every 10 years and a pandemicwill emerge within a 30 year time frame.

Reactive vaccination A two-dose reactive vaccination strategy is considered assuming that individuals are naïve to the pandemic virusand such reactive vaccines will be available after 6 months of the pandemic emergence.

Scenarios

Scenario 1 No pandemic occurs in the 30 year time frame and people have been pre-emptive vaccinated annually.

Scenario 2 A pandemic occurs in the 30 year time frame with a virus that does not match pre-pandemic vaccines i.e. vaccineefficacy is 0%. Reactive vaccines will be developed and administrated.

Scenario 3 A pandemic occurs in the 30 year time frame with a virus that matches pre-emptive vaccines with vaccine efficacyof 30%.

Scenario 4 A pandemic occurs in the 30 year time frame with a virus that matches pre-emptive vaccines with vaccine efficacyof 75%.

Social distancing and antiviral interventions

Limited duration schoolclosure

School were closed for 8 weeks of time as follows: for a primary school the whole school was closed if 1 or morecases were detected in the school; in a high school only the class members of the affected class were isolated(sent home and isolated at home) if no more than 2 cases were diagnosed in a single class; however if there weremore than 2 cases diagnosed in the entire high school the school was closed.

Sustained school closure All schools were closed simultaneously after a pandemic emerges.

Community contact reduction Community contact reduction was modelled by assuming that on days when the intervention was in effect allindividuals made 50% fewer random community contacts. This intervention is also modelled for either limitedduration or sustained.

Antiviral treatment andprophylaxis

Antiviral drugs used for treatment and prophylaxis of household members (of a symptomatic case) were combinedwith vaccination and social distancing interventions. It was assumed that 50% of symptomatic individuals would beidentified for antiviral treatment and/or prophylaxis, and that treatment and prophylaxis would occur 24 hours afterthe appearance of symptoms.


Pandemic characteristicsTwo plausible pandemic scenarios (moderate and severe)were defined based on the transmissibility and severitycharacteristics of past pandemics. The transmissibility of apandemic was defined in terms of its basic reproductionnumber R0 and associated illness attack rate. The severityof a pandemic was defined in terms of the case fatality rate(CFR).In this study, a moderate pandemic was defined as

having a reproduction number R0 of 1.9 and a CFR of0.25%. The reproduction number of the 1957 and 1968pandemics has been estimated to be in the range 1.5 and2.0 [43-45] with CFRs between 0.03% and 0.26% [2,3]. Asevere pandemic was defined with a reproduction num-ber R0 of 2.7 and a CFR of 1.5%, to reflect the estimatedtransmissibility and severity characteristics of the 1918pandemic, thought to have had a reproduction numberbetween 2.0 and 2.9 [44,46-48] with an estimated CFRbetween 0.74% and 1.8% [2-4].A proposed pandemic severity index from the US

Centers for Disease Control (CDC) uses the CFR for

categorising pandemic severity [49]. Severity categorieswere proposed from category 1, with a CFR < 0.1% tocategory 5, with a CFR > =2.0%. Our modelled scenariosfit into category 2 for moderate pandemics and category 4for severe pandemics.

Pre-emptive vaccinationA two-dose pre-emptive vaccination strategy was consid-ered, as trials of candidate pre-pandemic H5N1 vaccinesindicate the need for a two-dose regime to induce immun-ity [17-27]. It was assumed that pre-emptive vaccinationwould be an ongoing process and that the pre-pandemicvaccine would be updated and the population re-vaccinatedevery ten years (with additional vaccination in interveningyears at the population birth rate to maintain coverage).This is to ensure that the pre-pandemic vaccine reflects theinfluenza strains circulating in wildfowl and poultrypopulations that give most concern from a zoonoticperspective – at the current time these might be H5N1and H7N9, for example. One possibility for delivery ofthe pre-pandemic vaccine might be to include it as part


of a seasonal influenza vaccination program in eachpre-pandemic re-vaccination year, although it may notbe possible to achieve high coverage using this strategy. Itwas assumed that the average time between pandemicswas 30 years (based on the spacing of the previous 4 in-fluenza pandemics), so that on average 3 pre-pandemicvaccines would be developed and deployed for eachpandemic. Sensitivity analyses to both the 10 year vac-cination renewal period and 30 year pandemic returnperiod were conducted.Full (100%) vaccination coverage was assumed as a base-

line value; however since high levels of influenza vaccin-ation coverage may be difficult to achieve without theimmanent threat of a highly pathogenic pandemic, sensi-tivity analyses of different levels of vaccination coveragewere conducted. We make no assumption on the rate ofpre-emptive vaccination, other than that a certain level ofcoverage (100% for baseline scenarios, 10%-100% for sen-sitivity analyses) is achieved before the next pandemicoccurs.Trials of candidate pre-pandemic vaccines for the H5N1

influenza virus have shown seroconversion rates (definedas having a fourfold neutralizing seroconversion rate)between 60 and 90 per cent [18,20-22]. However, these pre-pandemic vaccines may not be closely matched to an emer-gent influenza strain or may offer only limited cross-strainprotection within the virus subtype, thus an efficacy of 30%is assumed. For completeness, a high efficacy pre-pandemicvaccine which closely matches the virus subtype with 75%efficacy was also considered; a scenario where the pre-pandemic vaccine was mismatched (having 0% efficacy)was also considered, in which case we assumed that a vac-cine would be developed reactively with a 6-month delay(see below). For individuals aged 65 years or older, we as-sume that vaccine efficacy will be 80% of the value for thoseyounger than 65. Thus for our assumed pre-pandemic vac-cine, we assume a VE of 24% for those 65 years and older,and a VE of 60% for a well-matched vaccine. A review ofstudies examining antibody responses to influenza vaccin-ation in older individuals found that seroconversion rateswere from 60% to 80% those of younger people, dependingon the influenza strain [50]. In the case of a well matchedvaccine (which we assume as having VE of 75% for youngerpeople), we assume a VE of 60%, which is consistent with atrial that found a VE of 58% for ages 60 and older for awell-matched seasonal vaccine [51].

Reactive vaccinationA two-dose reactive vaccination strategy was also con-sidered, assuming that individuals are naïve to a futureinfluenza strain and that a two-dose vaccine is essentialto achieve immunity [22,23]. During the H1N1 2009pandemic the first supplies of a suitable vaccine becameavailable after 5–6 months following the appearance of

the new strain of H1N1 influenza. In this study, a6 months delay from the onset of the pandemic to theinitiation of a vaccine campaign is assumed, after whichthe population is vaccinated at a rate of 1% of the popu-lation per day. As for the pre-emptive vaccination strat-egy, full vaccination coverage was assumed.Availability of a highly effective vaccine is assumed: tri-

als of candidate vaccines for the H1N1 2009 pandemicinfluenza showed seroconversion rates of vaccines be-tween 82 and 92 per cent [52]. Recent studies estimatedthat the vaccine effectiveness of the H1N1 2009 pan-demic vaccines was between 72% and 87% [53-55]; vac-cines with an efficacy of 75% are therefore assumed(60% in individuals aged 65 years or more).It was assumed that vaccination was prioritised so that

age groups known to have higher transmission rateswould be vaccinated first. Previous modelling resultshave indicated that a transmitters-first vaccination strat-egy is more effective in reducing both attack and casefatality rates than a vulnerable-first approach whentimely and wide vaccination coverage is possible [28,56](note that the issue of vaccination targeting is addressedfurther in the Discussion section).

Vaccination scenariosThe reactive vaccination strategy described above is thebaseline against which pre-emptive vaccination strategieswere compared. Four plausible pre-emptive vaccinationstrategies were considered and are defined in Table 1.In scenario 1 an assumption that no pandemic occurs

in the 30 year time-frame is made. Thus, there is nopandemic related mortality (and no life years saved), andthe only cost incurred is that of pre-emptive vaccination.In scenario 2, it is assumed that a pandemic does occur

in the 30 year time-frame, and the emergent pandemicvirus subtype does not match the able pre-pandemicvaccine with which the population has been vaccinated.No protection would be conferred by the pre-emptivevaccine (i.e. a vaccine efficacy of 0%), and it is assumedthat a new reactive vaccine would be developed andused after a 6-month delay. The effectiveness of thisstrategy in saving lives is thus the same as the reactivevaccination strategy, but the costs are the combinedcosts of pre-emptive and reactive vaccination.In scenario 3 it is assumed that the virus subtype

matches the pre-pandemic vaccine previously used, suchthat the vaccine efficacy is 30%.For completeness, scenario 4 considers a pandemic strain

which happens to be particularly well matched to the pre-pandemic vaccine, giving a vaccine efficacy of 75%.

Social distancing and antiviral drug interventionsFor the reactive vaccination strategy and each of the pre-emptive vaccination scenarios, three different combinations


of social distancing and antiviral drug-based interventionswere considered. The first is vaccination alone (i.e. with noadditional social distancing or antiviral usage). The secondis a combination of social distancing (school closureand community contact reduction) and antiviral treat-ment and prophylaxis, which is assumed to operate for8 weeks prior to and overlapping with the peak of thepandemic. The third is the same combination of socialdistancing and antiviral usage, except that it is assumedto continue for a sustained period (until the time whena reactive vaccination programme would be complete).Full details of non-vaccination interventions are pre-sented in Additional file 1.Antiviral drug and school closure interventions (either

sustained or limited duration) were initiated when spe-cific threshold numbers of symptomatic individuals werediagnosed in the community, which triggered healthauthorities to activate the intervention response; detailedanalyses of these assumptions were presented in previ-ous modelling studies [37,38,57].Community contact reduction was modelled by as-

suming that on days when the intervention was in effectall individuals made 50% fewer random community con-tacts. It was assumed that 50% of symptomatic individualswould be identified for antiviral treatment, that householdmembers of treated individuals would receive antiviralprophylaxis, and that treatment and prophylaxis wouldoccur 24 hours after the appearance of symptoms.

Influenza transmission modelIn the simulation model, the location and activity ofeach individual in the community is calculated twice perday – during one of the daily cycles (representing thenight, evening and morning) individuals were assumedto be at home and made potentially infective contactwith all other member of their household. In the otherdaily cycle (representing day-time activities) individualswere either located in their household, in a workplace,or in a school classroom, and made contact with a subsetof individuals in the same location; random community-based contact also occurred during the day cycle. Fulldetails of the human movement and contact model can befound in Additional file 1 and also in [15,28].Infectious transmission occurred when an infectious

and susceptible individual came into contact during asimulation cycle. Following each contact, a new infectionstate for the susceptible individual (either to remain sus-ceptible or to become infected) was randomly chosen viaa Bernoulli trial [37]. Once infected an individual pro-gressed through a series of infection states according toa fixed timeline.The probability that a susceptible individual would be

become infectious after an infectious contact was calcu-lated according to the following transmission function,

which takes into account the disease infectivity of theinfectious individual Ii and the susceptibility of suscep-tible individual Is at the time of contact.

Ptrans Ii; Isð Þ ¼ β� Inf Iið Þ � Susc Isð Þ � AVF Ii; Isð Þ� Vaccine Isð Þ

Each factor contributing to the transmission probability(underlying transmissibility β, time-varying transmissibilityInf(Ii), age-based susceptibility Susc(Is), antiviral effect-iveness AVF(Ii, Is), and vaccine effectiveness Vaccine(Is))is described in detail in [15,28] and Additional file 1.The transmission probability coefficient β, capturing theinfectivity of the virus strain, was chosen to give unmiti-gated epidemics with a specific basic reproductionnumber: R0 = 1.9 and 2.7 have been used in this study tocapture transmission characteristics for moderate andsevere influenza pandemics respectively.Epidemics were initiated by introducing one randomly

located infected seed individual into the population eachday, for the duration of the epidemic. All simulationswere repeated 40 times with random numbers control-ling the outcome of stochastic events (the locality ofseeded infected individuals and the outcome of infectivecontact events) and the results were averaged. Analysisof this simulation model has shown that the 40-runmean attack rate is highly unlikely (95% confidence) todiffer by more than 1.2% of the mean attack rate of amuch larger set of experiment repeats. In addition, anestimation of confidence interval (95% confidence) forthe cost effectiveness ratio for moderate and severe pan-demics is presented in Additional File 1: Figure A1.3).For all results we find that the 95% confidence intervalsare much smaller than the reported differences in costeffectiveness ratios for different strategies.We used C++ programing language to create individual-

based simulation model and Haskell functional program-ming for scripting purposes. The data produced by thesimulator were fed into an economical model developedusing MS excel to generate figures and tables.

Health outcomesIn order to calculate costs arising from lost productivityand from hospitalisation of ill individuals, an estimate ofindividual health outcomes (symptomatic illness, hospi-talisation, ICU admission and death) is required for eachsimulated scenario. The number symptomatic illnesseswas calculated from simulation results, which generatedage-specific attack rates for a pandemic with a particulartransmissibility and intervention scenario, assuming that32% of infections resulted in asymptomatic illness [58].Pandemic related mortality was then calculated usingthe number of simulated symptomatic cases and the(symptomatic) CFR according to the pandemic severity


(0.25% for a moderate pandemic and 1.5% for a severepandemic). As in prior cost effectiveness studies con-ducted using this simulation model [15,35-37], H1N12009 pandemic data from Western Australia were used toprovide a relationship between the CFR and numbersrequiring hospitalisation and ICU treatment. These dataindicated a non-ICU hospitalisation to fatality ratio of 32:1and an ICU admission to fatality ratio of 3:1 and align withthose derived by other studies, for example in [59].

Vaccination cost and economic analysisAs with the method presented in previous cost effective-ness studies using this simulation model [15,39-41], aneconomic model was used to translate the health out-come of each individual in the modelled population, asderived by the Albany simulation model and the healthoutcomes model, into the overall pandemic cost burden.The approach taken determines the total economic costto society incurred during a future influenza pandemic.Total costs involve both direct healthcare costs (e.g. thecost of medical attention due to a GP visit, or for hospi-talisation) and costs due to productivity loss due to ill-ness and social distancing interventions. Pharmaceuticalcosts (i.e. costs related to antiviral drugs and vaccines)are also estimated. Productivity losses due to death werenot considered in the analyses leading to the main re-sults of this study, however alternative analyses whichinclude productivity losses resulting from pandemicdeaths were conducted and these additional results arepresented in Additional file 1.The calculation of vaccination costs for reactive vac-

cination and pre-emptive vaccination differ from eachother. Reactive vaccines were those which were devel-oped, produced and deployed once the virus subtype ofan emergent pandemic was identified, and costs associ-ated with development, manufacture and administrationwere a once-off cost. In contrast, pre-pandemic vaccineswere assumed to be manufactured and administered asan on-going process, with an annual recurrent cost. Itwas assumed that the pre-pandemic vaccine would beupdated every ten years, to ensure that it would reflectthe most likely new emergent strain. For each pre-pandemic vaccine developed, it was assumed that thewhole population would be vaccinated in the year of de-velopment, and that each year thereafter full coveragewould be maintained by vaccinating infants (accordingto the average Australian birth rate). Conservative alter-nate assumptions that pre-pandemic vaccines wouldneed to be updated and deployed more often (e.g. every5 or 7 years) were examined in a sensitivity analysis.All costs are reported in 2012 US dollars using con-

sumer price index adjustments. 2012 US dollar valuesare used to make the results readily convertible to a widerange of developed countries. A full description of the

economic analysis, including cost data used in establish-ing the overall cost of each pandemic scenario is givenin Additional file 1. All future costs and life years werediscounted at 3% annually [60]. The time horizon forthis analysis is the period up to and including the nextinfluenza pandemic, which was assumed to be on aver-age 30 years based on the spacing of the previous 4influenza pandemics.

Cost effectivenessThe cost effectiveness of a given intervention strategy ispresented in terms of cost per Life Years Saved (LYS).The numerator used in this cost effectiveness ratio wasderived from the total cost arising from a given interven-tion being applied to the whole community. The denom-inator was calculated as the difference between years oflife lost for an unmitigated pandemic and a pandemicwith the intervention applied. Years of life lost data werederived for each simulation from the ages and life ex-pectancies of the individuals who died as a result of thepandemic.The cost effectiveness of each intervention is presented

as a cost in dollars ($) per LYS per person. This was de-rived by establishing the total cost for a particular inter-vention strategy and then dividing it by the population ofthe Albany model, approximately 30,000 individuals, soallowing the results to be applied to a population of anysize.

ResultsKey findingsAssuming that an influenza pandemic occurs every 30 yearson average, and that pre-pandemic vaccines are at least30% effective, a pre-emptive vaccination strategy is moreeffective, less costly, and more cost effective in terms of costper life year saved than a reactive vaccination strategy, forall severities and combinations of interventions simulatedin this study.Conversely, if the pre-pandemic vaccines developed

are ineffective (failing to provide protection due to strainmismatch), the policy of pre-emptive vaccination is lesscost effective than the reactive vaccination strategy, sincewe assume that a reactive vaccine will also be developedand deployed under these circumstances.The use of social distancing (SD) and antiviral (AV) in-

terventions in addition to either reactive or pre-emptivevaccination is always more effective than the vaccinationalone. The cost effectiveness of combining additional SDand AV with vaccination depends on the efficacy of pre-pandemic vaccines, the duration of social distancing,and the severity of pandemics.In the following sections we examine the relative effect-

iveness, total cost and cost effectiveness of pre-emptiveand reactive vaccination for all four pandemic scenarios.


Since the actual scenario that will occur in a future pan-demic is uncertain, we then examine the sensitivity of therelative cost effectiveness of pre-emptive and reactive vac-cination policies to key assumptions.The main outcomes of this study, namely the effective-

ness (LYS), the total cost, and cost effectiveness as thecost per life year saved ratio of each pre-emptive and re-active vaccination strategy, with and without social dis-tancing and antiviral interventions, for each modelledscenario, and for both moderate and severe pandemicsare presented in Tables 2, 3 and 4 respectively. An esti-mation of confidence interval for the cost effectiveness

Table 2 Effectiveness (Life years saved per 10,000) of vaccina(R = 1.9 and CFR = 0.25%) and severe pandemics (R = 2.7 and

Moderate pande

Scenarios SD and AV inter

Scenario 1: no pandemic none

Scenario 2: pandemic with virus which does not matchpre-emptive vaccine i.e. vaccine efficacy of 0%

none

8 weeks SD + AV

Sustained SD + AV

Scenario 3: pandemic with virus which matchespre-emptive vaccine with efficacy of 30%

none

8 weeks SD + AV

Sustained SD + AV


none

8 weeks SD + AV

Sustained SD + AV

Extreme pandem

Scenarios SD and AV inter

Scenario 1: no pandemic none


none

8 weeks SD + AV

Sustained SD + AV


none

8 weeks SD + AV

Sustained SD + AV


none

8 weeks SD + AV

Sustained SD + AV

*indicates life years saved (LYS) value not statistically significantly different from zer

ratio for moderate and severe pandemics for scenario 2is presented in Additional File 1. Figure 2 presents a costeffectiveness plane for scenario 2, scenario 3 and sce-nario 4, which plots each vaccination strategy with andwithout social distancing and antiviral interventionshorizontally according to the costs (compared to nointervention), and vertically according to the effective-ness (life years saved per 10,000 population). Figure 3presents cost breakdown of different cost componentssuch as health care cost, social distancing and illnesscost, antiviral cost and vaccination cost for all simulatedpandemic scenarios and interventions.

tion strategies for moderate pandemicsCFR = 1.5%)

mics (R = 1.9 and CFR = 0.25%)

vention Life years saved (LYS) per 10,000 population

Pre-emptive vaccination Reactive vaccination

- -

Pre-emptive + Reactive vaccination Reactive vaccination

4* 4*

253 253

421 421


312 4*

415 253

429 421


453 4*

460 253

462 421

ics (R = 2.7 and CFR = 1.5%)

vention Total cost ($) per 10,000 population


- -


9* 9*

1,219 1,219

2,720 2,720


1,680 9*

2,719 1,219

3,134 2,720


3,499 9*

3,507 1,219

3,510 2,720

o due to stochastic simulation variation.

Table 3 Total cost of vaccination strategies for moderate pandemics (R = 1.9 and CFR = 0.25%) and severe pandemics(R = 2.7 and CFR = 1.5%)

Moderate pandemics (R = 1.9 and CFR = 0.25%)

Scenarios SD and AV intervention Total cost ($) per 10,000 population


Scenario 1: no pandemic none $2,430,000 $0



none $6,510,000 $4,080,000

8 weeks SD + AV $7,510,000 $5,080,000

Sustained SD + AV $10,460,000 $8,030,000

Unmitigated pandemic cost (no intervention) $3,200,000



none $4,250,000 $4,080,000

8 weeks SD + AV $5,020,000 $5,080,000

Sustained SD + AV $6,820,000 $8,030,000


Pre-emptive Vaccination Reactive vaccination


none $3,630,000 $4,080,000

8 weeks SD + AV $4,800,000 $5,080,000

Sustained SD + AV $6,310,000 $8,030,000


Extreme pandemics (R = 2.7 and CFR = 1.5%)

Scenarios SD and AV intervention Total cost ($) per 10,000 population


Scenario 1: no pandemic none $2,430,000 $0



none $17,480,000 $15,050,000

8 weeks SD + AV $15,000,000 $12,570,000

Sustained SD + AV $13,430,000 $11,000,000




none $10,190,000 $15,050,000

8 weeks SD + AV $8,420,000 $12,570,000

Sustained SD + AV $8,810,000 $11,000,000




none $3,760,000 $15,050,000

8 weeks SD + AV $5,100,000 $12,570,000

Sustained SD + AV $6,970,000 $11,000,000



Reactive vaccinationA comprehensive cost effectiveness analysis of reactivevaccination including 6 months of development delay hasbeen conducted in [15]. Reactive vaccination alone is anineffective strategy in saving lives and not cost effective(see Tables 2 and 4, reactive vaccination column andFigure 2), since vaccination occurs well after the pandemic

peak and has essentially no effect. The addition of at least8 weeks social distancing and antiviral interventions to thereactive vaccination greatly increases effectiveness andcost effectiveness (see Tables 2 and 4). Sustained SD plusAV further improve effectiveness and cost effectiveness forboth moderate and severe pandemics (see Tables 2 and 4,reactive vaccination). For severe pandemics, the addition

Table 4 Cost effectiveness (cost per life years saved) of vaccination strategies for moderate pandemics(R = 1.9 and CFR = 0.25%) and severe pandemics (R = 2.7 and CFR = 1.5%)

Moderate pandemics (R = 1.9 and CFR = 0.25%)

Scenarios SD and AV intervention Cost ($) per LYS


Scenario 1: no pandemic occurs none - -



none $1,627,500 $1,020,000

8 weeks SD + AV $29,684 $20,079

Sustained SD + AV $24,846 $19,074



none $13,622 $1,020,000

8 weeks SD + AV $12,096 $20,079

Sustained SD + AV $15,897 $19,074

Pre-emptivevaccination Reactive vaccination


none $8,013 $1,020,000

8 weeks SD + AV $10,435 $20,079

Sustained SD + AV $13,658 $19,074

Extreme pandemics (R = 2.7 and CFR = 1.5%)

Scenarios SD and AV intervention Cost ($) per LYS


Scenario 1: no pandemic occurs none - -



none $1,942,222 $1,672,222

8 weeks SD + AV $12,305 $10,312

Sustained SD + AV $4,938 $4,044



none $6,066 $1,672,222

8 weeks SD + AV $3,097 $10,312




one $1,075 $1,672,222

8 weeks SD + AV $1,454 $10,312



of sustained SD and AV to reactive vaccination improvesthe relative cost effectiveness by decreasing total pandemiccosts and increasing the number of life year saved. Formoderate pandemics, it is the most cost effective strategybut it also increases the total pandemic costs (see Tables 3and 4, reactive vaccination).

Scenario 1: no pandemic occursOne of the main assumptions of this study is that a fu-ture pandemic will occur in a 30 year time frame. If nopandemic occurs, obviously there is no need for reactivevaccines. Therefore, reactive vaccination policy where novaccine is actually developed costs nothing, while the

pre-emptive vaccination policy that costs approximately$2.5 million per 10,000 population (see Table 3 – sce-nario 1). This scenario is has been included essentially toquantify the direct cost incurred by pre-emptive vac-cination. Note that in this scenario cost effectiveness isundefined, as no lives are lost due to the pandemic orsaved due to interventions.

Scenario 2: pandemic virus strain does not match pre-pandemic vaccines i.e. vaccine efficacy of 0%In this scenario, it was assumed that the ineffectivenessof the pre-pandemic vaccine would be quickly recog-nised, and a reactive vaccine would be developed and

Figure 2 Cost effectiveness plane for intervention strategies for different scenarios. Intervention strategies are labelled as R: Reactivevaccination, P: Pre-emptive vaccination, R + LSD: Reactive vaccination + Limited Social Distancing (8 weeks of SC + CCR) + AV, P + LSD: Pre-emptivevaccination + Limited Social Distancing + AV, R + SSD: Reactive vaccination + Sustained Social Distancing (Sustained SC + CCR) + AV, P + SSD:Pre-emptive vaccination + Sustained Social Distancing + AV, SC – School Closure, CCR – Community Contact Reduction, AV – Antiviral treatmentfor cases and prophylaxis for their household members, LSD – Limited Social Distancing, SSD – Sustained Social Distancing. All LSD and SSDinterventions integrate the use of antivirals (AV).


deployed to mitigate the effect of the pandemics. Theeffectiveness and cost effectiveness results, for bothmoderate and severe pandemics, are as for the reactivestrategy, except that under the pre-emptive vaccinationpolicy the additional cost of the pre-pandemic vaccineis paid, making the pre-emptive policy less cost effect-ive (see Tables 2, 3 and 4 for scenario 2). For example,the relative cost effectiveness ratio of the pre-emptive

Figure 3 Cost breakdowns to different costing components. Interventi

policy is approximately $25,000/LYS whereas $19,000/LYS for the reactive policy, for moderate pandemics(see Table 4 for scenario 2). Although these results areunsurprising, quantification of the additional cost ofhaving to develop and use both vaccines allows us todetermine how the cost-effectiveness of a pre-emptivevaccination policy degrades with increasing probabilityof vaccine failure.

ons strategies are labelled as Figure 2.


Scenario 3: pre-pandemic vaccine matches pandemicvirus with efficacy of 30%Moderate pandemicsFor moderate pandemics, the results suggested that pre-emptive vaccination strategies with or without social dis-tancing and antivirals were more effective, less costlyand more cost effective compared to their reactive vac-cination counterparts (see Tables 2, 3, 4 and Figure 2 –centre panel, green-diamond and green-cross). This isbecause for moderate pandemics, the cohort of immuneindividuals created by vaccination reduces the effectivereproduction of the pandemic and allows limited dur-ation social distancing and antiviral measures to behighly effective.The most cost effective strategy was pre-emptive vac-

cination coupled with 8 weeks SD plus AV. This resultedthe cost effectiveness ratio of $12,000 per LYS while sav-ing approximately 415 life years per 10,000 population,whereas the reactive counterpart resulted $20,000 perLYS while saving approximately 253 life years per 10,000population (see Table 4 for moderate pandemics). Themost effective strategy was the pre-emptive vaccinationcombined with sustained SD plus AV, which savedapproximately 430 life years per 10,000 population. Thiswas the most costly pre-emptive strategy however, cost-ing approximately $6.8 million per 10,000 population,because its sustained social distancing interventionsincurred additional costs (see Figure 3 – centre panel,moderate pandemics) while saving relatively few add-itional lives.

Severe pandemicsFor severe pandemics, the results indicated that themost effective and cost effective strategies included pre-emptive vaccination combined with sustained SD plusAV. This strategy is shown in Tables 2, 4 and Figure 2 –centre panel (labelled as P + SSD red-diamond for severepandemics).This strategy was highly cost effective due to the fact

that with a high CFR, prevention of infection translatedinto prevented hospitalisation, ICU usage and death. Thus,all interventions reduced both life years lost and costs,rendering tem highly cost effective. This can be seen inFigure 3 – scenario 3 for severe pandemics, which pre-sents the break-down of cost components for each inter-vention strategy and scenario. Due to this combinedeffect, this strategy could save approximately 3100 lifeyears per 10,000 population, which was the highest num-ber of life years saved, and cost approximately $2,800 perLYS, which was the lowest cost effectiveness ratio amongthe modelled interventions (see Tables 2 and 4).The next most cost effective strategy was that which

combined pre-emptive vaccination with 8 weeks SD plusAV (see Table 4 and Figure 2 centre panel – P + LSD red

diamond). When compared to the reactive counterpart,the combined pre-emptive vaccination with 8 weeks SDplus AV is approximately 55% more effective and approxi-mately 30% less costly (see Tables 2 and 3 for scenario 3).This is explained by the fact that in the case of the reactivevaccination policy the pandemic resurges once social dis-tancing interventions are relaxed, whereas the usage ofpre-pandemic vaccines creates a significant proportion ofimmune individuals which prevents the pandemic resur-gence, resulting in a significant reduction in mortality andtotal cost.

Scenario 4: pre-pandemic vaccine matches pandemicvirus with efficacy of 75%Moderate pandemicsUnder this scenario, the pre-emptive vaccination policywas highly effective and cost effective compared to thereactive vaccination policy. The most cost effective strat-egy was purely pre-emptive vaccination, resulting a costper life year saved ratio of $8000. The addition of socialdistancing and antivirals incurred total costs resulting in$10,400/LYS for 8 weeks social distancing plus antiviralsand $13,700/LYS for the sustained social distancing plusantivirals respectively. The additional social distancing andantiviral costs of these combined strategies incurs prod-uctivity losses due to social distancing interventions, with-out preventing many further symptomatic infections ordeaths. This is apparent in Figure 3 – right panel, whichshows the cost breakdown of different cost components.For severe pandemics, the results for scenario 4 follow

the same pattern as for moderate pandemics, but aremore pronounced, since the highly effective interventionsaves both lives and cost.

Sensitivity analysisSensitivity analyses were conducted for key model param-eters to determine how alternative assumptions about pre-emptive vaccination affected the relative cost effectivenesspre-emptive and reactive vaccination strategies.We specifically examined pre-emptive vaccination cover-

age, the time between pre-pandemic vaccine renewals, theexpected time between pandemics and the chance of pre-pandemic vaccine mismatch. Significant findings are sum-marised below. We further performed sensitivity analysesfor lower/higher weekly wages found in EU [61] and in-cluding/excluding productivity losses due to illness. Thesesensitivity analyses are presented in Additional File 1. Fur-ther sensitivity analysis of key parameters related to reactivevaccination can be found in details in the supplementary in-formation of [15].

Vaccination coverageWe assume that for a moderate or severe pandemic,having a CFR of 0.25% or greater, 100% or near 100%


coverage is plausible for a reactive vaccination strategy.For pre-emptive vaccination, where there is no deadlypandemic underway to focus the attention of the public, itmay not be possible to achieve a high level of vaccinationcoverage. We therefore extended our cost effectivenessanalysis to vary pre-emptive vaccination coverage from10% to 100%, for all scenarios and interventions consid-ered in this study (see Figure 4).For moderate pandemics, pre-emptive vaccination with

20% coverage coupled with 8 weeks social distancing andantiviral interventions resulted in better cost effectivenessthan the most cost effective reactive vaccination strategy;achieving 50% pre-emptive vaccination coverage was suffi-cient for near optimal (within 10%) cost effectiveness (seeFigure 4 – left panel).For severe pandemics, pre-emptive vaccination with at

least 30% coverage together with sustained social distan-cing and antivirals resulted in better cost effectivenessthan the most cost effective reactive vaccination strategy,and 60% pre-emptive vaccination coverage was near op-timal cost effectiveness (see Figure 4 – right panel).

Vaccine renewal frequencyDepending on the rate of emergence of new potential pan-demic virus strains and the degree of cross-strain reactivitythat can be achieved with a pre-emptive vaccine, vaccinesmay have to be renewed more frequently than the 10 yearbaseline value assumed in the main results. We alterna-tively assumed that the pre-pandemic vaccine would berenewed every 7 or 5 years, incurring new vaccine develop-ment and administration costs. The analyses suggested thatmore frequent renewal makes pre-emptive vaccination lesscost effective; however for both moderate and severe pan-demics the most cost effective strategy is still pre-emptivevaccination coupled social distancing, even if vaccine re-newal time is reduced to 5 years (see Figure 5).

Pandemic frequencyIf pandemics occurred at a lower rate than the averageof one every 30 years assumed in the main results, pre-emptive vaccination would become less cost effective,since the pre-emptive vaccine renewal program wouldresult in more vaccines being developed per pandemic.We examined the sensitivity of the results to pandemicfrequency by calculating costs and cost effectivenessassuming average time between pandemics ranging from10 to 100 years (see Figure 6). It was found that pre-emptive vaccination (coupled with social distancing)remained the most cost-effective strategy for pandemicreturn times up to 80 years.

Pre-pandemic vaccine mismatchThe vaccine effectiveness of a pre-pandemic vaccine is amajor unknown: we have examined three scenarios of

0%, 30% and 75% effectiveness. However without prob-abilities attached to these scenarios, a quantitative as-sessment of cost-effectiveness in the face of uncertaintycannot be made. We conducted a simple analysis byassuming that VE would be 0% effective with someprobability p, and 30% effective with probability (1-p). Inother words, there is a probability of p of complete vac-cine failure. Figure 7 shows the expected (in the tech-nical sense) cost effectiveness as a function of p. Thisshows that the most cost effective pre-emptive vaccin-ation strategy (including social distancing and antivirals)remains more cost effective than reactive strategies witha vaccine failure probability of 40% or lower.

DiscussionThe possible emergence of a highly pathogenic influenzapandemic presents a challenge to public health planning.Vaccination with an effective vaccine is an ideal solution;however such a vaccine would take at least 6 monthsto develop, test, produce and distribute with currentproven techniques, by which time the pandemic is likelyto have reached most parts of the world. Antiviral neur-aminidase drugs and social distancing measures of suffi-cient rigor would be capable of supressing transmission,however these could be extremely expensive and pos-sibly unsustainable [15,29,40,62].A pre-pandemic vaccine, developed to provide cross-

clade protection and administered prior to the pandemic,has the potential to address this problem. However thisstrategy comes with its own challenges: it is not certainwhat degree of vaccine efficacy that can be achievedagainst an unknown future influenza strain, it may be diffi-cult to achieve high levels of vaccination coverage, thepre-pandemic would need to be periodically renewed andre-administered.This modelling study has compared the relative effect-

iveness and cost effectiveness of reactive and pre-emptivevaccination strategies. In order to make a meaningfulcomparison, both strategies have been made as plausibleas possible: the reactive strategy assumes a six monthdelay to vaccine availability; and the pre-emptive strategyassumes a lower vaccine efficacy and ongoing vaccinerenewal costs. All other assumptions that are not relatedto vaccination have been kept the same for both vaccin-ation policies, including the inclusion of various combina-tions social distancing and antiviral measures.For moderate and severe influenza pandemics a policy

of pre-pandemic vaccination combined with social dis-tancing and antiviral interventions was found to be moreeffective and cost effective than a policy of reactive vac-cination, provided that at least 30% of the populationcan be pre-vaccinated. While a pre-pandemic vaccine ispresumed to have a lower efficacy than a matched reactivevaccine, pre-vaccination creates a cohort of individuals

Figure 4 Impact of vaccination coverage on the cost effectiveness. Interventions strategies are labelled as Figure 2.


who are immune at the start of the pandemics. Formoderate pandemics this reduces the effective repro-duction of the pandemic and allows limited durationsocial distancing and antiviral measures to be highlyeffective. For severe pandemics, the infection avoidedby pre-vaccination reduces both the number of liveslost and, due to avoided medical costs, the total cost ofthe pandemic.The most effective and cost effective strategies for

both reactive and pre-emptive vaccination involved thecombination of vaccination with social distancing and

Figure 5 Impact of vaccine renewal frequency on the cost effectivene

antiviral measures. For reactive vaccination, sustainedsocial distancing and antiviral measures (lasting for6 months) was always most cost effective, since withoutsome early intervention the pandemic peaked and sub-sided prior to the vaccination programme.For pre-emptive vaccination the optimally cost-effective

duration of social distancing depended on the pandemicseverity: for severe pandemics sustained social distancingwas the most cost effective, while for moderate pandemics8 weeks of social distancing was more cost effective thansustained social distancing. In this case the presence of

ss. Interventions strategies are labelled as Figure 2.

Figure 6 Impact of pandemic frequency on the cost effectiveness. Interventions strategies are labelled as Figure 2.


immune individuals, and the use of optimally timed [38]social distancing for 2 months was highly effective at redu-cing attack rate and mortality, to such a degree that add-itional months of social distancing did not significantlyfurther reduce the attack rate.If it transpired that the pre-pandemic vaccine was highly

effective (VE 70%), the cohort of immune individuals was

Figure 7 Impact of pre-pandemic vaccine mismatch on the cost effec

sufficient to prevent the pandemic, and pre-vaccinationalone was then highly cost effective.

Sensitivity analysisPre-emptive vaccination is an untried strategy, and whilewe have chosen plausible parameter values to representthis strategy in our model, it is possible that there are

tiveness. Interventions strategies are labelled as Figure 2.


other plausible assumptions that might lead to a differ-ent assessment of cost effectiveness. In fact, our sensitiv-ity analysis indicates that the results are quite robust tovariation in the most important of these assumptions.Unlike a reactive vaccination program, which we as-

sume will be able to achieve high vaccination coverageduring a highly pathogenic pandemic, it is possible thatvaccination coverage of a pre-pandemic vaccinationprogramme will not be able to achieve high levels ofcoverage. We found that pre-emptive strategies remainedmore cost effective compared to corresponding reactivestrategies for pre-emptive vaccination coverage levelsdown to 30% (while reactive strategies remain at 100%coverage). Pre-vaccination also remained more cost-effective than reactive vaccination if pandemics wereassumed to occur on average every 80 years, rather thanthe baseline assumption of 30 years; or if pandemic vac-cines needed to be renewed every 5 years rather thanthe baseline assumption of every decade. Our sensitivityanalysis also showed that if a policy of pre-emptive vac-cination is backed up by reactive vaccination in the caseof pre-pandemic vaccine failure (i.e. our scenario 2results), the pre-emptive policy can tolerate a vaccinefailure rate of 30% and remain competitive with areactive-only policy.Our economic analysis was conducted from a societal

perspective, and includes costs due to productivitylosses stemming from illness and social distancing in-terventions. Analysing the results from a health-careprovider perspective and omitting these costs, we findthat the most cost effective strategy as judged by thelowest cost per LYS ratio remains the same (resultsshown in Additional File 1).As with all modelling studies, the numerical results

presented in this study depend on assumptions andmodel parameters which are not known exactly. However,the results and trends discussed highlight the relative costeffectiveness of alternative intervention strategies – in thiscase, pre-emptive versus reactive vaccination strategies –that differ only on intervention parameters, and for whichall the other model parameters are the same. These resultsshould be robust to plausible variations in other modelparameter values, as such variations in model parametervalues will influence both simulation outcomes same way.

Related researchReactive vaccination strategies including vaccine devel-opment and deployment delay has been well studied in[15,16,32,35]. However, while no studies to our know-ledge have explicitly compared the cost effectivenessof pre-emptive and delayed reactive strategies, severalstudies have considered strategies involving vaccin-ation prior to or at the beginning of a pandemic with avaccine of limited efficacy.

The study of Newall et al. [31] explicitly modelledpharmaceutical interventions for a severe pandemic,explicitly taking into account a 6-month delay before theuse of a matched vaccine. This included scenarios wherea lower efficacy stockpiled pre-pandemic vaccine wasused at the beginning of the pandemic in addition to amatched vaccine after the delay. Although differences inassumptions mean that the study of Newall et al. is notdirectly comparable to the current study, it did show thatsufficient coverage with a low-efficacy pre-pandemicvaccine could effectively mitigate the impact of a severepandemic, and like the current study it concluded thatcost effectiveness depended upon pandemic severity,pandemic frequency, vaccination coverage, and prob-ability of pre-pandemic vaccine strain mismatch.Several studies have included pre-vaccination in a

cost-effectiveness analysis of pandemic influenza inter-ventions [16,29,30,34]. While these studies did not com-pare pre-vaccination to vaccination strategies with ahigh-efficacy vaccine and a long delay, they did considerthe use of limited efficacy vaccines at the beginning of apandemic. Each of these studies found that with suffi-cient coverage, which depended upon vaccine efficacyand the use complementary social distancing and anti-viral measures, limited efficacy vaccines could cost-effectively mitigate a pandemic.

Limitations and further researchWe recognise a number of limitations of the currentstudy that suggest avenues for further research. Whileour univariate sensitivity analyses examined the effect ofkey unknowns, a probabilistic analysis could estimatecost-effectiveness while taking into account all plausiblecombination of unknowns. Such a multivariate analysisshould define a (joint) probability distribution over pan-demic transmissibility and severity, pandemic frequency,pre-pandemic vaccine efficacy, and actual vaccinationcoverage achieved; the expected effectiveness and costeffectiveness of intervention strategies could then beestimated by sampling from this distribution.We have assumed that pre-pandemic vaccination

would occur prior to a pandemic, perhaps by the inclu-sion of a pre-pandemic vaccine in a seasonal influenzavaccine. As noted previously, it may be difficult toachieve high levels of coverage in this way, given therelatively low uptake of influenza vaccines [63]. An alter-native strategy might be to manufacture and keep a largestockpile of pre-pandemic vaccine, and to vaccinate atthe appearance of a highly pathogenic pandemic. Thismay enable higher vaccination coverage, but would incuradditional costs as vaccines would have to be stored andcontinually renewed due to limited shelf life. This strategyhas been assessed in [30,34,64], but performing this ana-lysis in the framework of the current study would allow a


pre-pandemic stockpiled strategy it to be compared to thepre-emptive and reactive strategies.The current study focuses on using mass vaccination to

achieve a population-level reduction in transmission, withconsequent reduction in attack rates, serious illness anddeath. As such we have assumed that when less than 100%vaccine coverage occurs, a transmitters-first strategy isadopted were vaccine is targeted towards those age groupsmost responsible for transmission (children and youngadults). Previous modelling studies have indicated that analternative vulnerable first strategy targeting groups at riskof serious illness and death due to influenza can be morecost-effective than the transmitters first strategy if vaccin-ation is delayed relative to the start of the pandemic[28,35,56]. Since our reactive strategy is delayed, this mightseem to indicate that we should simulate vulnerable-first.However our reactive strategy incorporates additional non-vaccination interventions during the delay period whichslow or arrest pandemic spread, which is not the case inthe other studies examining delayed vaccination. Since ourprevious study addressing the combination of delayed vac-cination and additional interim measures showed thattransmitters first vaccination was effective, we have retainedthis assumption in the current study [15]. We also note thatthe optimal choice between a transmitters-first and vulner-able-first can also depend upon population demographicsand the presence of pre-existing immunity [35].We have considered influenza pandemics of severity

category 2 and up on the CDC scale (i.e. with CFR >=0.25%). This corresponds to the 1918, 1956, and 1968pandemics, but not the 2009 pandemic. This focus wasmotivated by previous studies that indicated that react-ive vaccination was unlikely to be cost effective [15,31].Finally, the demographic, economic and health care

assumptions of our simulation model are based onAustralian population characteristics. However, the re-sults of the studies that use our simulation model [36]are consistent with a variety of other individual-basedsimulation models at a variety of scales (e.g. small com-munity [65-67], country [68,69]), in as far as comparablepandemics and intervention strategies are being evalu-ated. We thus believe that the model is broadly repre-sentative of developed world cities: sensitivity analysesincluded in this study showed that the total cost ofa pandemic varies directly with income levels, sinceillness-related productivity losses make up a substantialpart of the pandemic cost, but that the pattern of rela-tive cost/LYS ratio between strategies is insensitive to awide range of income levels.

ConclusionThis study has quantified the effectiveness and costeffectiveness of plausible and comparable pre-emptiveand reactive vaccination strategies, with and without

accompanying social distancing and antiviral interven-tions. We found that provided a sufficient level of vac-cination coverage can be achieved (approximately 30%in the case of a severe pandemic), pre-pandemic vaccin-ation strategies have the potential to be more effectiveand cost-effective and reactive strategies where a vaccineis developed and deployed after the start of a pandemic.One clear risk of a pre-pandemic strategy is that if afuture pandemic were to be of a subtype completelymismatched with respect to the pre-pandemic vaccine,the upfront cost of development and production of thepre-pandemic vaccine would be wasted. However, if theprobability of vaccine failure can be kept sufficientlylow, a policy of pre-vaccination with back-up reactivevaccination in case of vaccine failure can still be morecost-effective than a policy of reactive vaccination alone.Our analysis includes the assumption that pre-pandemicvaccines would be continually renewed in order to trackthe strains most likely to present a pandemic influenzathreat.Another key finding is that non-vaccination interven-

tions, including antiviral treatment and prophylaxis, andsocial distancing measures such as school closure, arecomplementary to vaccination, improving the effective-ness and cost effectiveness of both pre-emptive andreactive vaccination.Reactive vaccination strategies exist which are as

effective at mortality reduction as pre-emptive strat-egies, though less cost effective. However these strat-egies require vaccination to be preceded by the use ofrigorous social distancing interventions, including strictschool closure and community contact reduction, and thiswould need to be sustained for many months, resulting ina high economic cost and great social disruption.

Additional file

Additional file 1: A model-based economic analysis of pre-pandemicinfluenza vaccination cost effectiveness.

Competing interestsThe authors declare that they have no competing interests.

Authors’ contributionsGM, JK and NH were responsible for the conception and design of thesimulation experiments. NH and JK were responsible for simulation andeconomic model implementation. NH conducted simulation experiments. Allauthors were involved in the analysis of simulation results and writing themanuscripts. All authors read and approved the final manuscript.

AcknowledgementsThis study was funded by a project grant to GM from the National Healthand Medical Research Council of Australia (number 10004415). The fundershad no role in study design, data collection or analysis, decision to publish,or preparation of the manuscript.

Received: 11 December 2013 Accepted: 6 May 2014Published: 16 May 2014

http://www.biomedcentral.com/content/supplementary/1471-2334-14-266-S1.docx


References1. Dawood FS, Iuliano AD, Reed C, Meltzer MI, Shay DK, Cheng PY,

Bandaranayake D, Breiman RF, Brooks WA, Buchy P: Estimated globalmortality associated with the first 12 months of 2009 pandemicinfluenza A H1N1 virus circulation: a modelling study. Lancet Infect Dis2012, 12(9):687–695.

2. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K:Pandemic versus epidemic influenza mortality: a pattern of changingage distribution. J Infect Dis 1998, 178(1):53–60.

3. Glezen WP: Emerging infections: pandemic influenza. Epidemiol Rev 1996,18(1):64–76.

4. Frost W: Statistics of influenza morbidity with special reference to certainfactors in case incidence and case fatality. Public Heath Rep 1920,35:584–597.

5. Phillip C: Nature outlook: influenza. Nature 2011, 480(7376 Suppl):S1–S15.6. Li F, Choi B, Sly T, Pak A: Finding the real case-fatality rate of H5N1 avian

influenza. J Epidemiol Community Health 2008, 62(6):555–559.7. Mei Z, Lu S, Wu X, Shao L, Hui Y, Wang J, Li T, Zhang H, Wang X, Yang F, Jin

J, Shang Y, Zhang W: Avian influenza A(H7N9) virus infections, Shanghai,China. Emerg Infect Dis 2013, doi:10.3201/eid1907.130523.

8. Yu H, Cowling BJ, Feng L, Lau EHY, Liao Q, Tsang TK, Peng Z, Wu P, Liu F,Fang VJ, Zhang H, Li M, Zeng L, Xu Z, Li Z, Luo H, Li Q, Feng Z, Cao B, YangW, Wu JT, Wang Y, Leung GM: Human infection with avian influenza AH7N9 virus: an assessment of clinical severity. Lancet 2013, 382:138–145.

9. Cowling BJ, Jin L, Lau EHY, Liao Q, Wu P, Jiang H, Tsang TK, Zheng J, FangVJ, Chang Z, Ni MY, Zhang Q, Ip DKM, Yu J, Li Y, Wang L, Tu W, Meng L,Wu JT, Luo H, Li Q, Shu Y, Li Z, Feng Z, Yang W, Wang Y, Leung GM, Yu H:Comparative epidemiology of human infections with avian influenzaA H7N9 and H5N1 viruses in China: a population-based study oflab-confirmed cases. Lancet 2013, 382:129–137.

10. Viboud C, Simonsen L: Timely estimates of influenza A H7N9 severity.(Commentary). Lancet 2013, 382:106–108.

11. Herfst S, Schrauwen EJA, Linster M, Chutinimitkul S, de Wit E, Munster VJ,Sorrell EM, Bestebroer TM, Burke DF, Smith DJ: Airborne transmission ofinfluenza A/H5N1 virus between ferrets. Science 2012, 336(6088):1534–1541.

12. Imai M, Watanabe T, Hatta M, Das SC, Ozawa M, Shinya K, Zhong G, HansonA, Katsura H, Watanabe S: Experimental adaptation of an influenza H5 HAconfers respiratory droplet transmission to a reassortant H5 HA/H1N1virus in ferrets. Nature 2012, 420:420–428.

13. Russell CA, Fonville JM, Brown AEX, Burke DF, Smith DL, James SL, Herfst S,van Boheemen S, Linster M, Schrauwen EJ: The potential for respiratorydroplet-transmissible A/H5N1 influenza virus to evolve in a mammalianhost. Science 2012, 336(6088):1541–1547.

14. WHO: Pandemic influenza vaccine manufacturing process and timeline. [http://www.who.int/csr/disease/swineflu/notes/h1n1_vaccine_20090806/en/]

15. Kelso J, Halder N, Milne G: Vaccination strategies for future influenzapandemics: a severity-based cost effectiveness analysis. BMC Infect Dis2013, 13:81.

16. Andradóttir S, Chiu W, Goldsman D, Lee ML, Tsui KL, Sander B, Fisman DN,Nizam A: Reactive strategies for containing developing outbreaks ofpandemic influenza. BMC Public Health 2011, 11(Suppl 1):S1.

17. Plosker GL: A/H5N1 prepandemic influenza vaccine (whole virion, verocell-derived, inactivated). Drugs 2012, 72(11):1543–1557.

18. van der Velden MV, Fritz R, Pöllabauer EM, Portsmouth D, Howard MK, KreilTR, Dvorak T, Fritsch S, Vesikari T, Diez-Domingo J, Richmond P, Lee BW,Kistner O, Ehrlich HJ, Barrett PN, Aichinger G: Safety and Immunogenicityof a Vero Cell Culture-Derived Whole-Virus H5N1 Influenza Vaccine in aPediatric Population. J Infect Dis 2013, 1:12–23.

19. Herbinger KH, von Sonnenburg F, Nothdurft HD, Perona P, Borkowski A,Fragapane E, Nicolay U, Clemens R: A phase II study of an investigationaltetravalent influenza vaccine formulation combining MF59®: adjuvanted,pre-pandemic, A/H5N1 vaccine and trivalent seasonal influenza vaccinein healthy adults. Hum Vaccin Immunother 2013, 10(1):92–99.

20. Czajka H, Unal S, Ulusoy S, Usluer G, Strus A, Sennaroglu E, Guzik J, TopeliIskit A, Dargiewicz A, Musial D, Caylan R, Dziduch J, Eskioglu E, Hasiec B,Cwinarowiczliwa E, Belli R, Abdel-Messih IA, Beygo J, Fragapane E: A phase II,randomised clinical trial to demonstrate the non-inferiority of low-doseMF59-adjuvanted pre-pandemicA/H5N1 influenza vaccine in adult andelderly subjects. J Prev Med Hyg 2012, 53(3):136–142.

21. Vesikari T, Forstén A, Herbinger KH, Cioppa GD, Beygo J, Borkowski A, GrothN, Bennati M, von Sonnenburg F: Safety and immunogenicity of an

MF59(®)-adjuvanted A/H5N1 pre-pandemic influenza vaccine in adultsand the elderly. Vaccine 2012, 30(7):1388–1396.

22. Leroux-Roels I, Borkowski A, Vanwolleghem T, Dramé M, Clement F, Hons E,Devaster J-M, Leroux-Roels G: Antigen sparing and cross-reactive immunitywith an adjuvanted rH5N1 prototype pandemic infl uenza vaccine: arandomised controlled trial. Lancet 2007, 370:580–589.

23. Leroux-Roels I, Bernhard R, Gérard P, Dramé M, Hanon E, Leroux-Roels G:Broad clade 2 cross-reactive immunity induced by an adjuvanted clade1 rH5N1 pandemic influenza vaccine. PLoS One 2008, 3(2):e1665.

24. Lin J, Zhang J, Dong X, Fang H, Chen J, Su N, Gao Q, Zhang Z, Liu Y, WangZ, Yang M, Sun R, Li C, Lin S, Ji M, Liu Y, Wang X, Wood J, Feng Z, Wang Y,Yin W: Safety and immunogenicity of an inactivated adjuvantedwholevirion influenza A (H5N1) vaccine: a phase I randomised controlledtrial. Lancet 2006, 368:991–997.

25. Gasparini R, Amicizia D, Lai PL, Panatto D: Aflunov(®): a prepandemicinfluenza vaccine. Expert Rev Vaccines 2012, 11(2):145–157.

26. Fragapane E, Gasparini R, Schioppa F, Laghi-Pasini F, Montomoli E,Banzhoff A: A Heterologous MF59-Adjuvanted H5N1 PrepandemicInfluenza Booster Vaccine Induces a Robust, Cross-Reactive ImmuneResponse in Adults and the Elderly. Clin Vaccine Immunol Nov,17(11):1817–1819.

27. Banzhoff A, Gasparini R, Laghi-Pasini F, Staniscia T, Durando P, Montomoli E,Capecchi P, di Giovanni P, Sticchi L, Gentile C, Hilbert A, Brauer V, Tilman S,Podda A: MF59-adjuvanted H5N1 vaccine induces immunologic memoryand heterotypic antibody responses in non-elderly and elderly adults.PLoS One 2009, 4(2):e4384.

28. Milne G, Kelso J, Kelly H: Strategies for mitigating an influenza pandemicwith pre-pandemic H5N1 vaccines. J R Soc Interface 2010, 7:573–586.

29. Sander B, Nizam A, Garrison LP Jr, Postma MJ, Halloran EM, Longini IM Jr:Economic evaluation of influenza pandemic mitigation strategies in theUnited States using a stochastic microsimulation transmission model.Value Health 2009, 12(2):226–233.

30. Lee VJ, Tok MY, Chow VT, Phua KH, Ooi EE, Tambyah PA, Chen MI:Economic analysis of pandemic influenza vaccination strategies inSingapore. PLoS One 2009, 4(9):e7108.

31. Newall AT, Wood JG, Oudin N, MacIntyre CR: Cost-effectiveness ofpharmaceutical –based pandemic influenza mitigation strategies. EmergInfect Dis 2010, 16(2):224–230.

32. Baguelin M, Hoek AJ, Jit M, Flasche S, White PJ, Edmunds WJ: Vaccinationagainst pandemic influenza A/H1N1v in England: a real-time economicevaluation. Vaccine 2010, 28(12):2370–2384.

33. Prosser LA, Lavelle TA, Fiore AE, Bridges CB, Reed C, Jain S, Dunham KM,Meltzer MI: Cost-effectiveness of 2009 pandemic influenza A(H1N1)vaccination in the United States. PLoS One 2011, 6(7):e22308.

34. Khazeni N, Hutton DW, Garber AM, Owens DK: Effectiveness andcost-effectiveness of expanded antiviral prophylaxis and adjuvantedvaccination strategies for an influenza A (H5N1) pandemic. Ann InternMed 2009, 151(12):840–853.

35. Lugnér AK, van Boven M, de Vries R, Postma MJ, Wallinga J: Costeffectiveness of vaccination against pandemic influenza in Europeancountries: mathematical modelling analysis. BMJ 2012, 345:e4445.

36. Milne GJ, Kelso JK, Kelly HA, Huband ST, McVernon J: A small communitymodel for the transmission of infectious diseases: comparison of schoolclosure as an intervention in individual-based models of an influenzapandemic. PLoS One 2008, 3(12):e4005.

37. Halder N, Kelso J, Milne G: Analysis of the effectiveness of interventions usedduring the 2009 H1N1 influenza pandemic. BMC Publ Health 2010, 10:168.

38. Halder N, Kelso J, Milne G: Developing guidelines for school closureinterventions to be used during a future influenza pandemic. BMC InfectDis 2010, 10(1):221.

39. Halder N, Kelso J, Milne G: Cost-effective strategies for mitigating a futureinfluenza pandemic with H1N1 2009 characteristics. PLoS One 2011,6(7):e22087.

40. Milne GJ, Halder N, Kelso JK: The cost effectiveness of pandemic influenzainterventions: a pandemic severity based analysis. PLoS One 2013,8(4):e61504.

41. Kelso JK, Halder N, Postma MJ, Milne GJ: Economic analysis of pandemicinfluenza mitigation strategies for five pandemic severity categories.BMC Public Health 2013, 13:211.

42. Milne GJ, Kelso JK, Halder N: Lower NI efficacy may significantly reducepandemic intervention effectiveness. BMJ Editorial 2012, 345:1.

http://www.who.int/csr/disease/swineflu/notes/h1n1_vaccine_20090806/en/

http://www.who.int/csr/disease/swineflu/notes/h1n1_vaccine_20090806/en/


43. Vynnycky E, Edmunds W: Analyses of the 1957 (Asian) influenzapandemic in the United Kingdom and the impact of school closures.Epidemiol Infect 2008, 136(2):166–179.

44. Viboud C, Tam T, Fleming D, Handel A, Miller MA, Simonsen L:Transmissibility and mortality impact of epidemic and pandemicinfluenza, with emphasis on the unusually deadly 1951 epidemic.Vaccine 2006, 24(44):6701–6707.

45. Longini I, Halloran M, Nizam A, Yang Y: Containing pandemic influenzawith antiviral agents. Am J Epidemiol 2004, 159(7):623–633.

46. Gani R, Hughes H, Fleming D, Griffin T, Medlock J, Leach S: Potential impactof antiviral drug use during influenza pandemic. Emerg Infect Dis 2005,1(9):1355–1362.

47. Mills CE, Robins JM, Lipsitch M: Transmissibility of 1918 pandemicinfluenza. Nature 2004, 432:904–906.

48. Mathews JD, McCaw CT, McBryde JMES, McCaw JM: A biological model forinfluenza transmission: pandemic planning implications of asymptomaticinfection and immunity. PLoS One 2007, 11:1–6.

49. US Centers for Disease Control: Interim pre-pandemic planning guidance:community strategy for pandemic influenza mitigation in the UnitedStates. 2007, URL [http://www.flu.gov/planning-preparedness/community/community_mitigation.pdf]

50. Goodwin K, Viboud C, Simonsen L: Antibody response to influenzavaccination in the elderly: a quantitative review. Vaccine 2006,24(8):1159–1169.

51. Govaert TM, Sprenger MJ, Dinant GJ, Aretz K, Masurel N, Knottnerus JA:Immune response to influenza vaccination of elderly people: Arandomized double-blind placebo-controlled trial. Vaccine 1994,12(13):1185–1189.

52. Greenberg ME, Lai MH, Hartel GF, Wichems CH, Gittleson C, Bennet J,Dawson G, Hu W, Leggio C, Washington D: Response to a monovalent2009 influenza A (H1N1) vaccine. N Engl J Med 2009, 361(25):2405–2413.

53. Breteler JK, Tam JS, Jit M, Ket JC, De Boer MR: Efficacy and effectiveness ofseasonal and pandemic A (H1N1) 2009 influenza vaccines in low andmiddle income countries: A systematic review and meta-analysis. Vaccine2013, 31(45):5168–5177.

54. Simpson CR, Ritchie LD, Robertson C, Sheikh A, McMenamin J: Effectivenessof H1N1 vaccine for the prevention of pandemic influenza in Scotland,UK: a retrospective observational cohort study. Lancet Infect Dis 2012,12(9):696–702.

55. Widgren K, Magnusson M, Hagstam P, Widerström M, Örtqvist Å, Einemo IM,Follin P, Lindblom A, Mäkitalo S, Wik O, Österlund A, Grünewald M, Uhnoo I,Linde A: Prevailing effectiveness of the 2009 influenza A(H1N1)pdm09vaccine during the 2010/11 season in Sweden. Euro Surveill 2013,18(15):20447.

56. Mylius SD, Hagenaars TJ, Lugner AK, Wallinga J: Optimal allocation ofpandemic influenza vaccine depends on age, risk and timing. Vaccine2008, 26:3742–3749.

57. Kelso JK, Halder N, Milne GJ: The impact of case diagnosis coverage anddiagnosis delays on the effectiveness of antiviral strategies in mitigatingpandemic influenza A/H1N1 2009. PLoS One 2010, 5(11):e13797.

58. Fox JP, Hall CE, Cooney MK, Foy HM: Influenza virus infections in Seattlefamilies, 1975–1979. Am J Epidemiol 1982, 116:212–227.

59. Cauchemez S, Carrat F, Viboud C, Valleron A, Boelle P: A Bayesian MCMCapproach to study transmission of influenza: application to householdlongitudinal data. Stat Med 2004, 23(22):3469–3487.

60. World Health Organisation: Making choices in health: WHO guide tocost-effectiveness analysis. In Geneva: World Health Organization; 2003.

61. United Nation Economic Commission for Europe (UNECE) statistical database.<www.unece.org> (Accessed 7th Marth, 2014).

62. Perlroth DJ, Glass RJ, Davey VJ, Cannon D, Garber AM, Owens DK: Healthoutcomes and costs of community mitigation strategies for an influenzapandemic in the United States. Clin Infect Dis 2010, 50(2):165–174.

63. Wu S, Yang P, Li H, Ma C, Zhang Y, Wang Q: Influenza vaccinationcoverage rates among adults before and after the 2009 influenzapandemic and the reasons for non-vaccination in Beijing China: across-sectional study. BMC Public Health 2013, 13:636.

64. Jennings LC, Monto AS, Chan PK, Szucs TD, Nicholson KG: Stockpilingprepandemic influenza vaccines: a new cornerstone of pandemicpreparedness plans. Lancet Infect Dis 2008, 8:650–658.

65. Glass RJ, Glass LM, Beyeler WE, Min HJ: Targeted social distancing designfor pandemic influenza. Emerg Infect Dis 2006, 12(11):1671–1681.

66. Wu JT, Riley S, Leung GM: Reducing the impact of the next influenzapandemic using household-based public health interventions. HongKong Med J 2009, 15(Suppl 9):38–41.

67. Gojovic MZ, Sander B, Fisman D, Krahn MD, Bauch CT: Modelling mitigationstrategies for pandemic (H1N1) 2009. CMAJ 2009, 181(10):673–680.

68. Ferguson NM, Cummings DA, Fraser C, Cajka JC, Cooley PC, Burke DS:Strategies for mitigating an influenza pandemic. Nature 2006,442(7101):448–452.

69. Ferguson NM, Cummings DA, Cauchemez S, Fraser C, Riley S, Meeyai A,Iamsirithaworn S, Burke DS: Strategies for containing an emerginginfluenza pandemic in Southeast Asia. Nature 2005, 437(7056):209–214.

doi:10.1186/1471-2334-14-266Cite this article as: Halder et al.: A model-based economic analysis of pre-pandemic influenza vaccination cost-effectiveness. BMC Infectious Diseases2014 14:266.

Submit your next manuscript to BioMed Centraland take full advantage of:

• Convenient online submission

• Thorough peer review

• No space constraints or color figure charges

• Immediate publication on acceptance

• Inclusion in PubMed, CAS, Scopus and Google Scholar

• Research which is freely available for redistribution

Submit your manuscript at www.biomedcentral.com/submit

http://www.flu.gov/planning-preparedness/community/community_mitigation.pdf

http://www.flu.gov/planning-preparedness/community/community_mitigation.pdf

http://www.unece.org

Documents

A model-based economic analysis of pre-pandemic …smr.csse.uwa.edu.au/pdf/2014 Halder A model-based economic analy… · A model-based economic analysis of pre-pandemic influenza